Applies to

• Cablecast SX and Flex Video servers*

* SD video servers do not support decoding of HD content mentioned in this article.

Answer

Cablecast video servers can play back a number of different file formats / video codecs / audio codecs combinations. In order to be compatible with the video server, these files must follow certain specific parameters. The purpose of this article is to expose the requirements specific to each file format / codec combination, and as a result provide users with the knowledge to fully utilize Cablecast servers' multi-format file playback*.

The article is organized in sections. The first section covers broadcast basic information common across all file formats. First we’ll cover video resolutions and audio specifications. Then we'll present each specific file format**, outlining details specific to each file wrapper. Finally we'll go over the restrictions imposed on specific video and audio codec content type, regarldess of the file format it's in.

* this article covers playback only

** file format and file wrapper are considered synonyms

Broadcast Video Resolutions

For the purpose of this article, broadcast video resolutions can be broken down into the following elements.

Widthnumber of pixels per lineHeightnumber of linesAspect Ratiothe aspect ratio of each pixel Frame Ratethe number of picture frames displayed per second Scan Modethe order in which each picture (field or frame) is displayed

Every file consumed by the video server should have its basic video resolution and audio details match one of the following resolutions’ parameters. This is the first step to achieve file compatibility. For example, if a file contains SD content, and the server is in the US, then the file’s video parameters should match the NTSC resolution below.

Note: Interlaced content rate is sometimes referred to as samples-per-second. In this case 29.97 frames-per-second is equal to 59.94 samples-per-second, each sample representing a video field. In order to simplify this article we’ll always refer to the number of frames-per-second. Cablecast video server also always works in frames-per-second units.

NTSC

width720 height*480aspect ratio4x3 or 16x9 frame rate29.97scan mode**

interlaced, second field top

progressive

* compressed video will span 480 lines, while uncompressed video usually spans 486 lines

** progressive scan content is supported

PAL

width720 height*576aspect ratio4x3 or 16x9 frame rate25scan mode***

interlaced, first field top

progressive

720p

width1280height720aspect ratio16x9frame rate59.94scan modeprogressive

1080i

width1920*height1080aspect ratio16x9frame rate29.97scan mode**

interlaced, first field top

progressive

* some encoders support a width of 1440 (MPEG2, H264/AVC).

** progressive scan content is supported, but is not to be confused with 1080p below

1080p

width1920height1080aspect ratio16x9frame rate59.94scan modeprogressive

Audio Specification

Audio specification is not affected by video resolution and as such is much easier to manage.

samples per second48000* (48Khz)

* with the exception of mpeg files, where 44100 samples-per-second (44.1Khz) is supported.

File Formats

The following sections go over each supported file format and how its content must be stored in order to guarantee compatibility with Cablecast SX video server.

The tables represent each supported combination of video and audio content inside the file wrapper.

MOV / MP4 files

MOV* and MP4 file support is equivalent in Cablecast SX video servers.

videoaudioH264 / AVC, AVC-IAACProResPCMDV/DVCAM, DVCPRO, DVCPRO50, DVCPRO HDPCMMPEG2MPEG1 layer 2MPEG2PCMAvid DVPCMAvid DnxHDPCM

* Playback does not rely on Quicktime.

PCM audio restrictions

Supports 2 to 16 channels of 16- or 24-bit (24 bit in a 32 bit container), 48 kHz PCM audio. Audio can be stored as stereo pairs, mono tracks, or one muxed track. Audio track packaging must be consistent across the entire file, meaning that all tracks must be exclusively mono, exclusively stereo, or exclusively multiplexed. Big endian or little endian, optimized for little endian.

Cablecast will only play out the first two audio channels.

AAC audio restrictions

Supports compressed audio data in AAC format, encoded using the Low Complexity (LC) profile at 48 kHz with 2, 4, or 8 audio channels. AAC audio is only supported with H.264/AVC video.

Cablecast will only play out the first two audio channels.

Closed captioning and AFD

Playback of closed caption and AFD is supported only if the video essence is H.264/AVC, and the information is stored according to the AVCHD, AVC-Intra, ANSI/SCTE 128, or ETSI EN 301 775 specification. For AVCHD and AVC-Intra, the information should be stored as user data unregistered SEI (Supplemental Enhancement Information) messages within the H.264/AVC stream’s essence. For ANSI/SCTE 128 and ETSI EN 301 775, the information should be stored as user data registered SEI messages within the H.264/AVC stream’s essence.

Note on AFD: If no AFD is decoded from the content file, Cablecast will inject the proper AFD code into the playback's ancillary data stream.

MPEG files

Video with multiplexed audio in a program stream or transport stream, packaged in an .mpg, .mpeg or .m2v file.

videoaudio

MPEG-2

MPEG-1 layer 2

Audio restrictions

Mpeg audio must be multiplexed with the video, at 44.1kHz or 48 kHz, 16 bits per sample.

Audio stream can contain 1, 2, 4, or 8 pairs of stereo audio tracks (2, 4, 8 or 16 channels of audio).

Closed captioning, AFD

Closed caption and AFD should be stored in the user data (metadata) of the compressed MPEG-2. Ancillary data stored in the MPEG-2 user data must follow the ATSC A/53 standard.

Note on AFD: If no AFD is decoded from the content file, Cablecast will inject the proper AFD code into the playback's ancillary data stream.

AVI files

videoaudioDV/DVCAM, DVCPRO, and DVCPRO50PCMDVCPRO HDPCMMPEG2-IPCMMPEG2-IBP (incl. HDV)PCM

Codecs

Following are details on how compressed content should be presented in order to maximize compatibility with Cablecast SX video servers, regardless of the enclosing file format.

Video codecs

H264

• Profile: Main, High, High 10, High 4:2:2
• SD: Level 2.2, 3. 3.1
• HD: Level 3.2, 4, 4.1, 4.2, 5, 5.1, 5.2
• AVC-Intra class 50
• AVC-Intra class 100
• Entropy: CABAC, CAVLC
• Max gop size (N) = 30*
• Frame rate mode must be constant
• Color primaries: BT 601, BT 709

* longer gop sizes will play but may result in parts of the playback not being realtime, specially around cut points inside sequences or when resuming a playback.

Mpeg

• Main profile @ Main level (4:2:0)
• Main profile @ High level (4:2:0)
• Main profile @ High level-1440 (4:2:0)
• 422 profile @ Main level
• 422 profile @ High level

DnxHD

• 36, 145, 220, 220x

ProRes

• ProRes 422
• ProRes 422 HQ
• ProRes 422 LT
• ProRes 422 Proxy

Audio codecs

PCM

• 48Khz with 2, 4, 8, 16 or 32 audio channels* at 24 bit (24 bit in 32 bit payload)**.
• 48Khz with 2 audio channels* at 16 bits.

* Cablecast SX video servers will only play the first 2 audio channels.

** 24 bit audio is put inside a 32 bit container, using the most significant bits (8-31). Some applications may see the audio as 32 bit.

AAC

Low Complexity profile at 48kHz with 1, 4, or 8 channels*.

* Cablecast SX video servers will only play the first 2 audio channels.

Troubleshooting

Active Lines

The active lines information below is information that can help figure out why some files are playing with jerky motion or excessive aliasing, specially in NTSC. As such, 3rd-party encoded files should follow the 'active lines' specifications below to ensure they can be played without any field inversion or temporal inversion artifacts (common mostly in MPEG2 480i NTSC footage).

The active lines information is here for quality purpose, failure to follow these guidelines won't affect other aspects of playback.

NTSC active lines

Although an uncompressed NTSC video frame has a Bottom Field First field dominance, MPEG/H264/ProRes video is expected to be stored inside the file with a Top Field First field dominance. This is due to the difference in height between the compressed content (480 lines) and the original content (486 lines). The compression process crops 5 lines off the top of the frame, and 1 line off the bottom before sending the video frame to the encoder, for a total of 6 lines. Since an odd number of lines is taken off at the top of the frame, the remaining content's field dominance is set to Top Field First while stored as compressed data. The exact line numbers are represented in the table below.

The decoder expects the MPEG/H264/ProRes content to follow the above guidelines. If the Mpeg/H264/ProRes content is detected as Top Field First, the decoder will position the decoded 480 lines starting at line 5, and will play the full-height-decoded-content using a Bottom Field First field dominance. Essentially reversing the effects of the 486 -> 480 lines encoding process described above.

If the compressed content's field dominance is incorrectly tagged, the positioning of the decoded 480 lines content inside the full height 486 lines video frame will result in field inversion artifacts.

CodecActive LinesUncompressed

F1: 21-263, F2: 283-525

p: 40-525

MPEG/H264/ProRes

480i F1: 23-262, F2: 286-525

480p 45-524

DVF1: 23-262, 285-524

Pal active lines

CodecActive LinesUncompressed

F1: 23-310, F2: 336-623

p: 45-620

MPEG/H264/ProRes

576i F1: 23-310, F2: 336-623

DVF1: 23-310, F2: 335-622

720p active lines

CodecActive LinesAll codecs26-745

1080i active lines

CodecActive LinesAll codecsF1: 21-560, F2:584-1123

1080p active lines

CodecActive LinesAll codecs42-1121